Canary tools turn a binary wrong-tool error into a typed diagnosis of the reasoning weakness, and a six-type taxonomy separates models by capability.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.AI 1years
2026 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Diagnosing Tool-Selection Reasoning in LLM Agents with Canary Tools
Canary tools turn a binary wrong-tool error into a typed diagnosis of the reasoning weakness, and a six-type taxonomy separates models by capability.